关键词: 混合流水车间调度, 灰狼算法, 平面镜成像, 逆序解码

Abstract: To solve the hybrid flow shop scheduling problem with the identical parallel machine,an improved grey wolf optimization was proposed to minimize the makespan.A mathematical model of the problem was established according its characteristics.A new calculation formula was proposed for the important parameter  in grey wolf algorithm to improve the global exploration capability at early stage and the local search ability at late.Meanwhile,as the wolves gathered toward decision-making wolves,for maintaining the diversity of wolves,a plane-mirror imaging learning strategy was proposed to avoid the algorithm falling into local optimization.Moreover,considering the unbalance of machine at each stage,the forward decoding and reverse decoding strategy was used to improve the probability of finding the optimal solution.The proposed algorithm was compared with other algorithms on the benchmark cases and an actual enterprise case,and the effectiveness and reliability of the algorithm were verified.

Key words: hybrid flow shop scheduling, grey wolf optimization, plane-mirror imaging, reverse decoding